Between 2004 and 2007, the (real) median household income in the United States increased 3.2 percent, as measured by data available from the Current Population Survey’s (CPS) Annual Social and Economic Supplement (ASEC). This statistic compares a cross-section of households in 2004 with another crosssection of households in 2007, but does not provide a picture of what happened to the same households over time. Medians, like those available from the CPS-ASEC, can conceal fluctuations in annual household income. In order to examine changes in the annual (real) income of the same households between 2004 and 2007, this report uses the longitudinal data available from the 2004 panel of the Survey of Income and Program Participation (SIPP). (Text Box: Household Income).¹

Income quintiles were constructed for 2004 and 2007 using data collected in the 2004 SIPP panel (Text Box: Constructing Income Quintiles). Longitudinal data make it possible to identify and analyze factors that may contribute to an increase or a decrease in household income (Text Box: What Makes the SIPP a Longitudinal Survey?).²

_______________
¹ The data for this report were collected between February 2004 and January 2008 from households interviewed in all 12 waves of the 2004 SIPP panel. The population represented (that is, the population universe) is the civilian noninstitutionalized population living in the United States. See the “Source of Data” section for more details. All income amounts are adjusted to reflect 2007 dollars, unless indicated otherwise.

² This report is an update of “Dynamics of Economic Well-Being: Fluctuations in the U.S. Income Distribution, 2001–2003,” Current Population Reports, P70-112, U.S. Census Bureau, November 2007; and “Dynamics of Economic Well-Being: Movements in the U.S. Income Distribution, 1996–1999,” Current Population Reports, P70-95, U.S. Census Bureau, July 2004. This report focuses on household income rather than family or individual income. Several notable studies that have similarly used household income to investigate mobility are D’Ambrosio, D., “Household Characteristics and the Distribution of Income in Italy,” Review of Income and Wealth, Series 47, No.1, 2001, pp. 43–64; and Jarvis, S. and S. P. Jenkins, “Low Income Dynamics in 1990s Britain,” Fiscal Studies, 1997, Vol. 18, No. 2, pp. 123–42.

• Among U.S. households, 68.1 percent in the top quintile and 67.4 percent in the bottom quintile were in these same quintiles in 2004 and 2007.³
• Among U.S. households, between 42.7 percent and 47.6 percent of households in the middle three quintiles were in these same quintiles in 2004 and 2007.
• Approximately 12.3 million U.S. households (11.5 percent) experienced changes in their annual income between 2004 and 2007 that resulted in their moving either up or down two or more quintiles in the income distribution.
• Of these 12.3 million households, approximately 2.3 million households in the bottom quintile and 2.0 million households in the second quintile experienced the largest percentage of gains in annual household income between 2004 and 2007.
• Of these 12.3 million households, 5.0 million households that started in the top and fourth quintiles experienced a decline of two or more quintiles between 2004 and 2007.
• Householders who had lower levels of education were more likely to remain in or move into a lower quintile compared with householders who had higher levels of education.
• Householders who were not married were more likely to remain in or move into a lower quintile compared with householders who were married.
• Younger householders (aged 15 to 24) were more likely than others to move down from the top and the fourth quintiles, while older householders (aged 65 and older) were most likely to remain in the bottom and the second quintiles.

_______________
³ The estimates in this report (which may be shown in text, figures, and tables) are based on responses from a sample of the population and may differ from the actual values because of sampling variability and other factors. As a result, apparent differences between the estimates for two or more groups may not be statistically significant. All comparative statements have undergone statistical testing and are significant at the 90 percent confidence level unless otherwise noted.